# Totally Spies S7 — three-axis framework for characterization ## The problem with our earlier writing We mixed up three things that operate on completely separate axes: 1. **Production technique** — *how* the animation is made 2. **Character design aesthetic** — *what* the characters look like 3. **Motion language** — *how* things move When we said "2D cutout" and "anime" in the same sentence as if they were both style descriptors, we created a false contradiction. They are not contradictory. They describe different axes entirely. --- ## Axis 1 — Production technique **What it describes:** the tools and workflow used to create the animation. **For S7:** Toon Boom Harmony rigged/cutout animation. Characters are built from hierarchical drawing layers (head, torso, arms, hands, mouth shapes, etc.), animated by setting keyframes on transforms, with Harmony interpolating in-betweens. This is a **production workflow decision** about efficiency, consistency, and cost. This axis has nothing to do with what the characters look like. You can make a show that looks like classic Disney using a cutout rig. You can make a show that looks like Totally Spies using a cutout rig. You can make a show that looks like Samurai Jack using a cutout rig. The rig is invisible in the output. --- ## Axis 2 — Character design aesthetic **What it describes:** the visual design language of the characters — their proportions, linework, expression conventions, color palette. **For S7 (and the whole franchise):** a deliberate East-West hybrid. Director Stéphane Berry said it directly: > "a melting between the American style, which associates action and comedy, > and Japanese design for the aesthetic environment and the emotions expressed > through the large eyes of the characters" Specifically, what makes the Totally Spies character design identifiably influenced by Japanese animation: - large eyes with distinct highlight shapes - relatively thin limbs and elongated proportions - specific facial structure (small nose, simplified mouth) - fashion and outfit focus (recognizable in anime shoujo genres) - expressive emotion conventions borrowed from anime (sweatdrop, speed lines, exaggerated reactions) What makes the overall show identifiably Western (and specifically French/ Canadian co-production): - action-comedy structure - Beverly Hills / spy thriller genre frame - broader physical comedy and slapstick - specific background design palette (more American-cartoon influenced) This axis has nothing to do with the production technique. The character designs were originally drawn for a **traditionally animated** show (S1–6). When S7 switched to rigged animation, the designers built rigs that reproduce the same design language. The aesthetic target did not change. The method of achieving it did. --- ## Axis 3 — Motion language **What it describes:** how much things move, what kinds of movement occur, and what motion conventions the show uses. **For S7:** computationally measured as: - ~65% of full-episode frame pairs are near-identical holds - ~10% are high-motion - motion is typically controlled, key-pose-driven - backgrounds are mostly static This axis is shaped by both the production technique (cutout rigs are inherently more economical with motion) and by the budget and scheduling constraints of a 26-episode TV series. It is also influenced by S1–6's legacy motion language, which S7 tries to retain. This axis has nothing to do with what the characters look like. A show can have anime-influenced character designs and be full-animation (actual anime productions often are). A show can have anime-influenced character designs and be very limited (Totally Spies S7 is). --- ## The correct description of S7 Using the three axes separately and precisely: **Production technique:** S7 is animated in Toon Boom Harmony using rigged/cutout puppets. This is the first season of the franchise to use this method. S1–6 were traditionally animated. **Character design aesthetic:** The franchise has an intentional East-West hybrid design language: anime-influenced character proportions and expression conventions (large eyes, elongated limbs, emotion iconography) in a Western action-comedy genre frame. S7 retains this design language from the earlier seasons. The aesthetic target did not change when the production method changed. **Motion language:** S7 operates in a hold-heavy mode (~65% near-identical frame pairs across full episodes) with real action content (~10% high-motion frame pairs). Motion is key-pose-driven and controlled, consistent with a TV-budget rigged-animation production. --- ## Why this matters for AI video implications These are three separate problems for any AI generation system: 1. **Production-technique problem:** the model needs to produce output that looks consistent with rig-based motion — smooth keyframe interpolation between poses rather than the subtle organic variation of frame-by-frame drawing. This is a temporal motion characteristic. 2. **Character-design-aesthetic problem:** the model needs to have learned the specific East-West hybrid design language. Generic Western cartoon training data will not transfer well. Generic anime training data will also not transfer well. The specific proportions, expression conventions, and color palette of Totally Spies are the target. 3. **Motion-language problem:** the model needs to maintain stable held frames for the majority of shots, produce controlled action for roughly 1 in 10 frame pairs, and not invent motion where none should exist. These are three distinct things to optimize. Conflating them — as our earlier writing did — produces an incoherent description and incomplete implications.